The use of a pre-defined coding template in constructivist simulation-based research

Abstract

Many constructivist methodologies and methods used in simulation-based research (SBR) involve coding of text, otherwise known as thematic analysis. This debate concerns whether, when thematically analysing textual data in SBR, codes should be pre-defined or derived from the data. Arguments in favour of using pre-defined codes included the grounding of a study within existing literature, building on that literature and explicitly defining the researchers’ initial knowledge and understanding. Failure to pre-define the codes may result in producing a brick to throw on the pile of SBR instead of constructively adding to existing knowledge. Arguments against the use of pre-defined codes included the risk of pre-forming conclusions at an early stage, and the potential to confine analysis, thus stifling the creation of new knowledge. SBR using pre-defined codes may create an ‘echo-chamber’ for pre-existing ideas and may result in ‘seeing only what we want to see’. Factors that may determine whether to pre-define codes include the desire for transferability of results to other contexts, and uniqueness of the topic. Researchers should be aware of the arguments in favour of each of the conflicting approaches, and make explicit their reasons for choosing one approach over another.